The Global Burden of Disease

The Treemap

Photo by National Cancer Institute on Unsplash

Without hair, a queen is still a queen…
— Prajakta Mhadnak

Global Burden of Disease Study, 2019

Ingest the data

disease by cause and death

# Load data
df <- read.csv("./archetypes/global-burden-of-disease/IHME-GBD_2019_DATA-GLOBAL-ALL-CAUSES.csv", header = TRUE, stringsAsFactors = TRUE)  # read text file 
df

df_taxonomy <- read.csv("./archetypes/global-burden-of-disease/IHME_GBD_2019_CAUSE_HIERARCHY_Y2020M11D25.csv", header = TRUE, stringsAsFactors = TRUE)  # read text file 
# df_taxonomy

Wrangle the data

Create hierarchy

df_wrangle <- df %>% filter(measure_name == "Deaths")
df_wrangle <- df_wrangle %>% filter(metric_name == "Number")
df_wrangle <- df_wrangle %>% select(cause_id, cause_name, val)

df_taxonomy_select <- df_taxonomy %>% select(Cause.ID, Cause.Name, Parent.ID, Parent.Name, Level)
df_taxonomy_l1 <- df_taxonomy_select %>% filter(Level == 1)
df_taxonomy_l2 <- df_taxonomy_select %>% filter(Level == 2)
df_taxonomy_l3 <- df_taxonomy_select %>% filter(Level == 3)
df_taxonomy_l4 <- df_taxonomy_select %>% filter(Level == 4)

df_tree_1_2 <- merge(df_taxonomy_l1, df_taxonomy_l2, by.x="Cause.ID", by.y="Parent.ID")
df_tree_1_2 <- df_tree_1_2 %>% select(Cause.ID, Cause.Name.x, Cause.ID.y, Cause.Name.y, )
colnames(df_tree_1_2) <- c("L1_ID", "L1_NAME", "L2_ID", "L2_NAME")
# df_tree_1_2

df_tree_2_3 <- merge(df_taxonomy_l2, df_taxonomy_l3, by.x="Cause.ID", by.y="Parent.ID")
df_tree_2_3 <- df_tree_2_3 %>% select(Cause.ID, Cause.Name.x, Cause.ID.y, Cause.Name.y, )
colnames(df_tree_2_3) <- c("L2_ID", "L2_NAME", "L3_ID", "L3_NAME")
# df_tree_2_3

df_tree_3_4 <- merge(df_taxonomy_l3, df_taxonomy_l4, by.x="Cause.ID", by.y="Parent.ID")
df_tree_3_4 <- df_tree_3_4 %>% select(Cause.ID, Cause.Name.x, Cause.ID.y, Cause.Name.y, )
colnames(df_tree_3_4) <- c("L3_ID", "L3_NAME", "L4_ID", "L4_NAME")
# df_tree_3_4

df_tree_all <- merge(df_tree_1_2, df_tree_2_3, by="L2_ID")
df_tree_all <- merge(df_tree_all, df_tree_3_4, by="L3_ID")
df_tree_all <- df_tree_all %>% select(L1_ID, L1_NAME, L2_ID, L2_NAME.x, L3_ID, L3_NAME.x, L4_ID, L4_NAME)
colnames(df_tree_all) <- c('L1_ID', 'L1_NAME', 'L2_ID', 'L2_NAME', 'L3_ID', 'L3_NAME', 'L4_ID', 'L4_NAME')
# df_tree_all

df_hierarchy <- merge(df_wrangle, df_tree_all, by.x="cause_id", by.y="L4_ID")
df_hierarchy

Analyze the data

Create a quartile column to use for color mapping

The within function is used for calculating new columns. The quantile function calculates the quartiles where 0:4/4 evaluates to c(0, 0.25, 0.50, 0.75, 1). Finally the cut function splits the data into the calculated quartiles. Casting the result to integers returns groups labeled as 1,2,3,4. The column is converted to a factor for use with a discrete color scale.

df_analysis <- within(df_hierarchy, quartile <- factor(as.integer(cut(val, quantile(val, probs=0:4/4), include.lowest=TRUE))))
df_analysis <- df_analysis %>% select(L1_NAME, L2_NAME, L3_NAME, L4_NAME, val, quartile)
df_analysis

The Layout

squarified with size and fill

# Layouts
# squarified" (the default), "scol", "srow" or "fixed"
# Simple
v1 <- ggplot(df_analysis, aes(area = val, fill = quartile)) +
  geom_treemap(color = "#ffffff") +
  scale_fill_brewer(palette = "Spectral") +
  theme_minimal() +
  theme(legend.position = "bottom")

girafe(ggobj = v1, width_svg = 1280/72, height_svg = 720/72,
       options = list(opts_sizing(rescale = TRUE, width = 1.0))
)

with labels

# Labeled
v2 <- ggplot(df_analysis, aes(area = val, fill = quartile, label = L4_NAME)) +
  geom_treemap(color = "#ffffff") +
  geom_treemap_text(family = "inconsolata", fontface = "italic", colour = "white", place = "centre", grow = TRUE) +
  scale_fill_brewer(palette = "Spectral") +
  theme_minimal() +
  theme(legend.position = "bottom")

girafe(ggobj = v2, width_svg = 1280/72, height_svg = 720/72,
       options = list(opts_sizing(rescale = TRUE, width = 1.0))
)

with labels and color palette

# Sub-grouping
v3 <- ggplot(df_analysis, aes(area = val, fill = quartile, label = L4_NAME,
                subgroup = L1_NAME)) +
  geom_treemap(color = "#ffffff") +
  geom_treemap_subgroup_border(colour = "white") +
  geom_treemap_subgroup_text(place = "centre", grow = T, alpha = 0.5, colour =
                               "black", family = "inconsolata", fontface = "italic", min.size = 0) +
  geom_treemap_text(family = "inconsolata", colour = "white", place = "topleft", reflow = T) +
  scale_fill_brewer(palette = "Spectral") +
  theme_minimal() +
  theme(legend.position = "bottom")

# v3

girafe(ggobj = v3, width_svg = 1280/72, height_svg = 720/72,
       options = list(opts_sizing(rescale = TRUE, width = 1.0))
)

as facets

# Sub-grouping
v4 <- ggplot(df_analysis, 
             aes(area = val, fill = quartile, label = L4_NAME, subgroup = L3_NAME)) +
  geom_treemap(color = '#ffffff') +
  facet_wrap( ~L2_NAME) +
  scale_fill_brewer(palette = "Spectral") +
  geom_treemap_subgroup_border(colour = "white") +
  geom_treemap_subgroup_text(place = "centre", grow = T, alpha = 0.5, colour = "black", family = "inconsolata", fontface = "italic", min.size = 0) +
  geom_treemap_text(family = "inconsolata", colour = "white", place = "topleft", reflow = T) +
  theme_minimal()+
  theme(legend.position = "bottom")

girafe(ggobj = v4, width_svg = 1280/72, height_svg = 1280/72,
       options = list(opts_sizing(rescale = TRUE, width = 1.0))
)

References

citations for narrative and data sources

Narrative and Data Sources:
Global Burden of Disease Collaborative Network.
Global Burden of Disease Study 2019 (GBD 2019) Results.
Seattle, United States: Institute for Health Metrics and Evaluation (IHME), 2020.
Available from http://ghdx.healthdata.org/gbd-results-tool.